It has recently been claimed that symptom reports cannot be used as indicators of disease because they are confounded by personality dispositions, e.g., the tendency to experience negative affect, a conclusion that would invalidate large bodies of data used to estimate health needs and to assess the effects of psychosocial factors, e.g., age, stressors, upon illness. Our conceptualization of symptoms as the result of the bodies immunological response to pathogens, predicts that both symptoms and immunological reactions will be less intense in older than younger persons and in depressed than non-depressed or anxious individuals. We also state that appraisal procedures intervene between symptoms and the decision that one is ill and in need of medical care, and that anxious and depressed individuals may fail to conclude they are ill and in need of health care if they attribute symptoms to their emotional states rather than illness. We also predict that depressed individuals will experience a greater number of vague, systemic symptoms rather than intense, disease specific symptoms, and that the accumulation of such episodes over a 3 year period will result in increase s in trait depression. These hypotheses will be tested in a 3 year longitudinal trial (LT) and two, smaller, embedded investigations. Four, 1&1/2 hour interviews will be conducted\at yearly intervals with 1,200 residents (ages 55 to 90) of a New Jersey retirement community. Measures of trait anxiety and depression, illness history, social support networks, life stress, preventive health behaviors, and reports of symptom episodes during the prior two weeks will be obtained at each interview. Four phone calls between the yearly intervals will tap and trak the initiation, interpretation, mode of coping and resolution of symptom episodes. A quasi experimental study will examine the symptomatic and antibody reactions to either tetanus or flu inoculation of 288 residents selected to vary on age and trait anxiety and depression. A second study examines the diurnal variation of symptom reports as a function of age and trait anxiety and depression. The complete data set should allow reasonably strong conclusions regarding the role of personality, age, preventive health actions and life stress on the formation and appraisal of symptoms.